SYNTHESIS OF POLYDEUTERATED a-DECALONES
April 5 , 1963 [COSTRIBUTIOS FROM THE
DEPARTMENT O F CHEMISTRY, STANFORD UNIVERSITY,
941
STANFORD, CALIF.]
Mass Spectrometry in Structural and Stereochemical Problems. XVIII. Fragmentation and Hydrogen Transfer Reactions after Electron Impact on a-Decalones. Synthesis of Polydeuterated a-Decalones2 BY E. LUpl;D,3H. BUDZIKIEWICZ, J. M. WILSONAND CARLDJERASSI RECEIVED NOVEMBER 10, 1962 Seventeen deuterated analogs of trans-decalone-1 (I), trans-9-methyldecalone-1 (XVITa) and cis-9-methyldecalone-1 ( X V I I b ) have been synthesized and their mass spectra investigated. By noting the occurrence or absence of shifts upon deuteration in any given mass spectral peak, plausible mechanisms could be proposed for most of the principal fragmentation processes. Furthermore, insight could be gained into the many hydrogen transfer reactions accompanying such fragmentation processes initiated by electron bombardment. Several of these processes are dependent upon the nature of the ring juncture and the stereochemical implications of these observations are pointed out.
As part of our program4 on the correlation of mass spectral fragmentation patterns and structural features of polycyclic molecules, there have been studied the mass spectra of a variety of steroid ketones5-’ The over-all course of the principal fragmentation processes can usually be determined by “labeling” different portions of the molecule with substituents such as alkyl groups. However, in order to gain insight into the fragmentation mechanism and especially the hydrogen transfer reactions incident upon electron impact, i t is necessary to have available deuterated analogs and an extensive synthetic program along these lines is currently under way in our laboratory. At the same time, synthetic and mass spectral studies were also initiated with simpler bicyclic model ketones, especially since the past literature is virtually devoid of relevant information in the field of cyclic ketones. In the present article we record the mass spectral results derived from an examination of various deuterated analogs of adecalone. Synthesis of Deuterated a-Decalones.-The only deuterated analog derivable from trans-a-decalone (I) itself is the 2,2,9-&-analog 11, which can be obtained readily by equilibration with sodium in a mixture of deuteriomethanol and heavy water. Since i t was also necessary to label all of the other “non-activated” positions with deuterium, a more versatile precursor had to be selected and this proved to be the 4-hydroxyA6-octalone-1 (111), a convenient preparation of which was recently described by Ireland and Marshall.* Modified Wolff-Kishner reduction8 of I11 followed by acetylation provided Ao-octalol-1-acetate which successively was reduced‘O with N-trideuterio-ptoluenesulfonylhydrazide, saponified and oxidized to provide 6,7-d~-trans-decalone-l(VII). The initial step in the synthesis8 of the hydroxyoctalone I11 is the Diels-Alder addition“ of butadiene to p-benzoquinone. (1) Paper X V I I , hf. Plat, J . LeMen, M.-M. Janot, H. Budzikiewicz, J. M . Wilson, I,. J . Durham and C. Djerassi, Buli. soc. chim. F m n c e , 2237 (1962). (2) We are indebted t o the National Institutes of Health of t h e U. S. Public Health Service for financial support (grants No. CRTY-5061 and S o . A-4257). (3) Taken from part I1 of the Ph.D. dissertation of E. Lund, Stanford University, 1963, where the mass spectra of the deuterated analogs (Tables 1-3) are reproduced. (4) For review see C. Djerassi, Puve Apbl. Chem., 6 , No. 4 (1963). ( 5 ) H. Budzikiewicz and C. Djerassi, J . A m . Chem. Soc., 84, 1430 (1962). ( 6 ) C . Djerassi, J. M. Wilson, H. Budzikiewicz and J , W. Chamberlin, ibid., 84, 4544 (1962). (7) C. Iljerassi, H. Budzikiewicz and J. hi. Wilson, “Recent Applications of Mass Spectrometry in Steroid Chemistry,” Proc. Internat. Congress Hormonal Steroids, Milano, May, 1962, Academic Press, Inc., New York, S . Y., in press. ( 8 ) R . E. Ireland and J . A. Marshall, J . O r g . C h e m . . 2 1 , 1620 (1962). (9) This material may still contain some cis fused isomer, but purification was always effected a t the ketone stage to provide the Irans-decalones, (10) See R . S. Dewey and E. E. van Tamelen, J . A m . Chem. S o c . , 8 8 , 3729 (196 1). (11) As modified by I,. F. Fiescr, ibid., 70, 3165 (1948).
By utilizing commercially available d6-butadiene and otherwise following the Ireland-Marshall sequence,8 there was obtained the 5,$6,7,8,8-d6 analog of transA6-octalone-1 (VI), which upon catalytic deuteration12 afforded the desired 5 ,5j6,6,7,7,8,8-d8-trans-decalone-1 (VIII). Mass spectrometry demonstrated the presence of 89% of the d8-ketone VIII, contaminated with 8% of the d7- and 3% of the d6-species. In order to label positions 3, 4 and 10 in the oxygenated ring, trans-4-hydroxydecalone-1 (IX)I3 was transformed into its ethylenemercaptal X and desulfurized with deuterated w - 7 Raney nickel catalyst.I4 In our experience, this represents the least satisfactory method of introducing deuterium selectively and oxidation of the chromatographically homogeneous desulfurization product led to 4,4-d2-trans-decalone-1(XII), accompanied by substantial amounts of d-, ds- and non-
V, R
=
H lation of 2-n-butylthiomethylenedecalone-1 i i t h methyl iodide followed by removal of the butylthiomethylene blocking ~ a 2 1 group as described by Ireland and h l a r ~ h a l l 'provided mixture of cts-iXVIIb) and trans-(XVIIa) 9-methyldecalone-1, which was readily separated by gas phase chromatography using a Beckman Megachrom instrument with a 6-ft. diethylene glycol succinate column operating a t 180" with helium as the carrier gas. The component of 15-min. retention time was shown to be the cis-isomer XVIIb by conversion to the oxime of m.p. 114llRo (lit.33m.p. 114.5-115.5'), while the trans isomer (m.p. of oxime 141-142', lit.33 m.p. 141.5-142.5") exhibited a retention time of 18 min. Utilizing the identical procedure with trideuteriomethyl iodide, there were obtained the trans-(XVIII) and cts-(XVIII) 9trideuteriomethyldecalones of 929; purity (determined mass spectrometrically). The usual equilibration of trans-9-methyldecalone-1 (23'IIa) provided the 2,2-d2 derivative XIXa ( 9 5 k d ~ 5% , d l species), (31) T h e alternate procedure (L. F. Fieser, ibid., 76, 1945 (1934)) of adding immediately t h e boron trifluoride gave only a 30% yield of t h e desired mercaptal X accompanied by dehydrated material. (32) See C. Djerassi and J. Staunton, ibid., 89, 734 (1961), Table 11. (33) W. S. Johnson, ibid., 66, 1317 (1943).
April 5, 1963
ROTATORY DISPERSIONO F 3-METHYL- A N D 4-METHYLCYCLOHEPTANOSE
949
trans-(XXIIa) and cis-(XXIIb)-J,J, 10-d3-9-Methyldecalone-l.while similar treatment of the cis-decalone XVIIb led to the 2,2The identical procedure applied to XT'I led to the ketones X X I I a d 2 analog X I X b of 100% purity. , 327, and trans-iXXa) and cis-(XXb)-5.5.6.6.7.7.8.8-dR-9-Methvldeca- and X X I I b of the isotope composition: ds 6 1 7 ~ d2 dl 7%. lone-1.-A sample (88 mg.) of t h e &-cz-'decalone VI11 was transtrans-(XXIVa) and cis-(XXIVb)-6,7-d~-9-Methyldecalone-l.formed by the published procedurels to the butylthiomethylene The mixture (267 mg.) of trans-(XXIIIa) and cis-(XXIIIb) 9derivative, methylated, the blocking group removed and the trans and cis isomers separated by g a s phase chromatography. methyl-A6-octaloness was reduced with lithium aluminum hyMass spectrometry showed the presence of 9Yc d,, 40% d8,.23yb dride in ether solution (20 hr., room temperature), the resulting d7, 11% dg and less than 47q each of do, d l , d2, da, d r , and d j species. alcohol acetylated, reduced with N-trideuterio-p-toluenesulfonyl trans-(XXIa) and czs-(XX1b)-4,4-d2-9-Methyldecalone-1 .hydrazide (see preparation of VII), saponified, oxidized and the Methylationl6 of X I 1 and gas phase chromatographic separation mixture (over-all yield, 93 mg.) of trans-(XXIVa) and cisgave the required ketones X X I a and X X I b , the composition of ( X X I V b ) isomers separated by gas phase chromatography on a which was determined mass spectrometrically as: d o 13%, Craig succinate column, the products consisting of 57% dz, 337, d, and 27Yc do species. di 3 8 7 ~ d2 ~3 5 7 ~ d t ~147,.
[CONTRIBUTION FROM
THE
DEPARTMENTS OF CHEMISTRY,WASHINGTON UNIVERSITY,ST. LOUIS, Mo. ; POLYTECHNIC ISSTITUTE OF BROOKLYS,BROOKLYN, s.Y.; STANFORD UNIVERSITY, STANFORD, CALIF.]
Absolute Configuration and Optical Rotatory Dispersion of 3-Methglcycloheptanone and 4-Methylcycloheptanone BY CARL DJERASSI,'B. F. BURROWS,' c. G. OVERBERGER,~ T. TAKEKOSHI,~ c. D. GUTSCHE3AND
c. T. CHANG3
RECEIVEDNOVEMBER 19, 1962
+
Ring enlargement of ( )-3-rnethylcyclohexanone, a compound of reliably established absolute configuration, is shown t o yield ( +)-3-methylcycloheptanone and ( - )-4-methylcycloheptanone. The reassignment of the sign of rotation and the sign of the optical rotatory dispersion Cotton effect to the 3-methyl isomer elirninates the contradictions posed by the stereochemical outcome of the diazomethane ring enlargement of optically active 2-methylcyclohexanones and confirms the proposition t h a t the diazoalkane-carbonyl reaction proceeds with retention of configuration.
A recent communication directed to the stereochemistry of the diazomethane ring enlargement of optically active 2-methylcycl~hexanone~ required the conclusion, based on the data contained therein and on the existing data in the literature, that the diazoalkanecarbonyl reaction proceeds with inversion of configuration. On the basis of other facts concerning such reactions, however, this appeared to be inadmissible and to call for a reinvestigation of some of the earlier experiments. One of the several critical points in the argument focused on the previously reported synthesis of ( - ) -3-methylcycloheptanone by diazomethane ring enlargement of (+)-3-methyl~yclohexanone.~ the absolute configuration of which had been established by methods generally accepted as reliable.6 The work reported in the present communication demonstrates that the product of ring enlargement of (+)-3-methylcyclohexanone is actually the (+)-rotating rather than the (- )-rotating 3-methylcycloheptanone, the earlier reported5 negative rotation (and negative Cotton effect) having been due to contamination with the strongly levorotatory 4-methylcycloheptanone which is also produced in the ring enlargement reaction. Thus the contradictions posed by the stereochemical outcome of the diazomethane ring enlargement of %methylcyclohexanone are resolved, and i t is firmly established that the diazomethane-carbonyl reaction proceeds with retention of configuration. Ring enlargement of (+) -3-methylcyclohexanone was carried out by the in situ diazomethane method according to the procedure previously employed by Djerassi and K r a k ~ w e r , ~by , ' the ex situ diazomethane (1) T h e work a t Stanford University was supported by grant no. C R T Y 5061 from t h e National Cancer Institute, U. S. Public Health Service. (2) T h e work a t t h e Polytechnic Institute of Brooklyn was supported by Chas. Pfizer, C o . , Inc. (3) T h e work a t Washington University was supported by grant no. 628-A4 from t h e Petroleum Research Fund administered by t h e American Chemical Society and by grant no. G-21323 from t h e National Science Foundation. (4) C. D. Gutsche and C. T. Chang, J . A m . Chem. S O L . ,84,2263 (1962). ( 5 ) C. Djerassi and G. W. Krakower, ibid., 81,237 (1959). (6) A . Fredga, A y k i v . X e m i .Vfinerai. Geol., PIA, KO.32 (1847); E. J . Eisenbraun and S. M ,McElvain, J . A m . Chem. Soc., 11, 3383 (195.5) (7) T . J . DeBoer and H. J . Backer, Oug. S y w l h e s e s , 94, 24 (1954); G. W. Krakower, Ph.D. Thesis, Wayne State University, 1958,
methodY8and by the Demjanow-Tiffeneau methodg; in all cases the product consisted of approximately equal amounts of the 3- and 4-methylcycloheptanones. The two materials, however, are not easily separated, and it is this difficulty that obscured the earlier report5 concerning their optical rotations. By subjecting the crude mixture to distillation through a very efficient spinning band column a t a very high reflux ratio or by passing the crude mixture through a sufficiently long and efficient vapor phase chromatographic (v.P.c.) column i t is possible to obtain one fraction with [ a ] 64' and another fraction with [ a ]- 137". Although resembling each other in boiling point, refractive index and nuclear magnetic resonance spectra (n.m.r.), the two fractions showed significant differences in detail in the infrared spectrum, in the mass spectrum, in V.P.C. behavior and especially in optical rotatory dispersion (R.D.) (Fig. 1). Furthermore, derivatives of the two fractions showed them to be quite different compounds. Identification of the (+)-fraction as (+)-3methylcycloheptanone (strong positive Cotton effect) was based on a comparison with material previously obtained4 from the ring enlargement of 2-methylcyclohexanone; identification of the (-)-fraction as (-)-4-methylcycloheptanone (strong negative Cotton effect) was based on a comparison with material obtained from the ring enlargement of 4-methylcyclohexanone. Additional support for these structural assignments was adduced from the degradative reactions resulting in the conversion of the respective ketones to the corresponding dibasic acids followed by pyrolytic cyclization to a mixture of niethylcyclohexanones which could be separated by v.p.c. As anticipated, application of this reaction sequence to (+)-3methylcycloheptanone provided 2- and 3-methylcyclohexanone, while ( - ) -4 -methylcycloheptanone afforded 3- and 4-methylcyclohexanone uncontaminated with 2methylcyclohexanone.
+
(8) C . 1). Gutsche, "Organic Reactions," Vol. V I I I , John Wiley and Sons, Inc., Kew York, N . Y . , 1954, p. 364. (9) P. A. S. Smith and D. R . Baer, "Organic Reactions," Vol. 11. John Wiley and Sons, Inc., New York, N. Y., 1960, p. 157.